Skip to content

Qwen3.8-Flash-Next on any consumer hardware: one-click install for Windows / Linux. Strata inference engine, OpenAI/Anthropic API on localhost, optional image input.

C++

metal: fp8 conversion modes, the round-to-nearest float helpers, and an mma.h placeholder

Niko1221e46b149ae558
diff --git a/metal/include/metal_compat/cuda_fp16.h b/metal/include/metal_compat/cuda_fp16.h
index e69de29bb..d00a77a8b 100644
--- a/metal/include/metal_compat/cuda_fp16.h
+++ b/metal/include/metal_compat/cuda_fp16.h
@@ -0,0 +1,7 @@
+
+// the round-to-nearest float helpers: on Apple Silicon the FP ops are already IEEE and un-fused here
+static inline float __fmul_rn(float a, float b) { return a * b; }
+static inline float __fadd_rn(float a, float b) { return a + b; }
+static inline float __fmaf_rn(float a, float b, float c) { return a * b + c; }
+static inline float __frcp_rn(float a) { return 1.0f / a; }
+static inline float __fsqrt_rn(float a) { return __builtin_sqrtf(a); }